- Run the shell script:
bash data_gen/data_gen_NLE/ReactionDiffusionEq/run_DarcyFlow2D.sh
which will in turn run the python script
data_gen/data_gen_NLE/ReactionDiffusionEq/reaction_diffusion_2D_multi_soluion_Hydra.py
- Update
data_gen/data_gen_NLE/config/config.yaml
to:
type: "ReacDiff" # 'advection'/'ReacDiff'/'burgers'/'CFD'
dim: 2
- Finally, run the data merge script:
python data_gen/data_gen_NLE/Data_Merge.py
# generate data and save as .npy array
cd PDEBench/pdebench/data_gen/data_gen_NLE/AdvectionEq
# Either generate a single file
CUDA_VISIBLE_DEVICES='2,3' python3 advection_multi_solution_Hydra.py +multi=beta1e0.yaml
# Or generate all files
bash run_trainset.sh
- Update
data_gen/data_gen_NLE/config/config.yaml
to:
type: "advection" # 'advection'/'ReacDiff'/'burgers'/'CFD'
dim: 1
savedir: "./save/advection"
# serialize to hdf5 by transforming npy file
cd ..
python Data_Merge.py
# generate data and save as .npy array
cd PDEBench/pdebench/data_gen/data_gen_NLE/BurgersEq/
# Either generate a single file
CUDA_VISIBLE_DEVICES='0,2' python3 burgers_multi_solution_Hydra.py +multi=1e-1.yaml
# Or generate all files
bash run_trainset.sh
- Update
data_gen/data_gen_NLE/config/config.yaml
to:
type: "burgers" # 'advection'/'ReacDiff'/'burgers'/'CFD'
dim: 1
savedir: "./save/burgers"
# serialize to hdf5 by transforming npy file
cd ..
python Data_Merge.py
# generate data and save as .npy array
cd PDEBench/pdebench/data_gen/data_gen_NLE/ReactionDiffusionEq/
# Either generate a single file
CUDA_VISIBLE_DEVICES='0,1' python3 reaction_diffusion_multi_solution_Hydra.py +multi=Rho2e0_Nu5e0.yaml
# Or generate all files
bash run_trainset.sh
- Update
data_gen/data_gen_NLE/config/config.yaml
to:
type: "ReacDiff" # 'advection'/'ReacDiff'/'burgers'/'CFD'
dim: 1
savedir: "./save/ReacDiff"
# serialize to hdf5 by transforming npy file
cd ..
python Data_Merge.py